Impact of Phosphorothioate Chirality on Double‐Stranded siRNAs: A Systematic Evaluation of Stereopure siRNA Designs

Abstract Oligonucleotides are important therapeutic approaches, as evidenced by recent clinical successes with antisense oligonucleotides (ASOs) and double‐stranded short interfering RNAs (siRNAs). Phosphorothioate (PS) modifications are a standard feature in the current generation of oligonucleotide therapeutics, but generate isomeric mixtures, leading to 2n isomers. All currently marketed therapeutic oligonucleotides (ASOs and siRNAs) are complex isomeric mixtures. Recent chemical methodologies for stereopure PS insertions have resulted in preliminary rules for ASOs, with multiple stereopure ASOs moving into clinical development. Although siRNAs have comparatively fewer PSs, the field has yet to embrace the idea of stereopure siRNAs. Herein, it has been investigated whether the individual isomers contribute equally to the in vivo activity of a representative siRNA. The results of a systematic evaluation of stereopure PS incorporation into antithrombin‐3 (AT3) siRNA are reported and demonstrate that individual PS isomers dramatically affect in vivo activity. A standard siRNA design with six PS insertions was investigated and it was found that only about 10 % of the 64 possible isomers were as efficacious as the stereorandom control. Based on this data, it can be concluded that G1R stereochemistry is critical, G2R is important, G21S is preferable, and G22 and P1/P2 tolerate both isomers. Surprisingly, the disproportionate loss of efficacy for most isomers does not translate into significant gain for the productive isomers, and thus, warrants further mechanistic studies.